I assume from his last name that Dan Choi, the Arabic-language specialist who is being dismissed because he's openly gay, is not of Middle Eastern extraction. But I take it that many Arabic-language specialists, including the gay ones, are.
So my question (which I'm sure is not original): Wouldn't the gay Arabic speakers be especially likely to be loyal to us, and hostile to Islamic fundamentalists? As between a gay Arabic speaker who has ties to, say, Iraq or Saudi Arabia, and the straight Arabic speaker, whom would you trust more to lack hidden sympathies with violent Islamic extremism?
Just to be clear -- I think it's entirely reasonable to worry about possible loyalty issues when it comes to selecting people for sensitive military or national security positions (whether high level or low), especially when the candidates are culturally, ethnically, or religiously linked to our enemies. For instance, I take it that the government should rightly have worried about this with regard to Russian immigrants like my family and me when the Cold War was still on, even though on balance the Russian immigrants were highly hostile to the country that they left. The normal antidiscrimination rules, important as they are, don't in my view apply equally to the military or national security, just as the normal sex discrimination laws don't apply equally to the military, and just as the normal free speech and search and seizure rules don't apply equally to the military or national security.
This doesn't itself tell us what the government should do based on those worries. I think the answer depends on many factors, including the degree of harm that comes to people because of these connections (e.g., internment vs. some extra investigation, exclusion from the military vs. placing Japanese-American soldiers predominantly on the European front during World War II, a tendency to exclude from a vast range of military jobs vs. a tendency to exclude from the most security-sensitive jobs). But it does suggest that we should avoid policies that end up excluding those people as to whom the risk of disloyalty seems especially low. And that's entirely independent of your stand on whether sexual orientation discrimination is wrong in principle, or counterproductive in general (setting aside our current enemies' extreme hostility to homosexuality).